UNIVERSITY OF PADUA FACULTY OF AGRICULTURE Department of Land and Agro-forestry Systems

Testo completo

(1)UNIVERSITY OF PADUA FACULTY OF AGRICULTURE Department of Land and Agro-forestry Systems. Erasmus Mundus Master Course in “Sustainable Forest and Nature Management” (SUFONAMA). ANALYSIS OF QUALITY DEVELOPMENT IN BROAD-LEAVED TREE-FARMING PLANTATIONS IN NORTHERN-EAST ITALY. Supervisor Prof. MARIO PIVIDORI Co-Supervisor Dr. CHIARA CANESIN Master student MARCO DAMIANI no. 606736-AB. Academic Year 2008– 2009.

(2) 2.

(3) Index. Summary ......................................................................................................................... 7 1. Introduction.............................................................................................................. 9 1.1 Background......................................................................................................... 9 1.2 Problem statement .............................................................................................. 9 1.3 Objectives and research questions .................................................................... 10 1.4 Structure of the thesis ....................................................................................... 11 2. Theoretical background ........................................................................................ 12 2.1 Definitions ......................................................................................................... 15 2.2 Theoretical approaches ...................................................................................... 16 3. Materials and methods .......................................................................................... 19 3.1 Research approach ............................................................................................ 19 3.2 Study area ......................................................................................................... 19 3.3 Sample plots...................................................................................................... 21 3.3.1 Instruments for measuring ............................................................................. 22 3.3.2 Data collection ............................................................................................... 22 3.4 Data analysis ..................................................................................................... 24 4. Results and Discussions ......................................................................................... 27 4.1 European walnut (Juglans regia L.) .................................................................. 28 4.2 Wild cherry (Prunus avium L.).......................................................................... 35 4.3 European ash (Fraxinus excelsior L.)................................................................ 42 5. Conclusions............................................................................................................. 48 References...................................................................................................................... 52 Annex 1 – Stem defects considered in this study ....................................................... 55 Annex 2 – Field paper used in the 2009 field data collection.................................... 56 Annex 3 – Additional graphs comparing walnut, cherry and ash results ............... 57 Annex 4 – Sample plots parameters ………….…………………...............................58. 3.

(4) List of figures. Figure 2.1: Timber assortments that can be obtained from tree-farming plantations Figure 2.2: Example of a mixed plantation with accessory species Figure 2.3: Types of tree farming plantations according to species composition Figure 2.4: Ideal central cylinder containing all defects Figure 3.1: Map of Gorizia province Figure 3.2: Stem straightness classes in Nosenzo’s classification Figure 3.3: Some stem defects considered in this study Figure 3.4. Tiny insect holes that affect only the bark of some European ash stems Figure 4.1: Mean DBH annual increment of walnut trees plotted against age Figure 4.2: Stem defects found on all walnut butt-logs Figure 4.3: Stem defects due to the presence of knots and branches on all walnut butt-logs Figure 4.4: Walnut butt-log grade distribution according to Nosenzo’s classification and Canesin’s classification Figure 4.5: Walnut butt-log grade distribution in both 2006 and 2009 according to Canesin’s classification Figure 4.6: Mean DBH annual increment of cherry trees plotted against age Figure 4.7: Stem defects found on all wild cherry stems. See Table 4.1 for the list of stem defects included in each group Figure 4.8: Stem defects due to the presence of knots and branches on all wild cherry butt-logs Figure 4.9: Wild cherry butt-log grade distribution according to Nosenzo’s classification and Canesin’s classification Figure 4.10: Wild cherry butt-log grade distribution in both 2006 and 2009 according to Canesin’s classification Figure 4.11: Mean DBH annual increment of cherry trees plotted against age Figure 4.12: Stem defects found on all European ash butt-logs Figure 4.13: Stem defects due to the presence of knots and branches on all European ash buttlogs Figure 4.14: European ash butt-log grade distribution according to Nosenzo’s classification and Canesin’s classification Figure 4.15: European ash butt-log grade distribution in both 2006 and 2009 according to Canesin’s classification. 4.

(5) List of tables. Table 2.1: Prices of walnut logs belonging to different grades in 2003 Table 3.1: List of the 21 plots sampled in 2009 with the respective species composition and municipality in which they are located Table 3.2: Comparison between the 2 classifications adopted in this study Table 4.1: Percentage of each stem defect found on all sampled trees of each principal species and grouped according to the origin and the stem feature they affect Table 4.2: Mean DBH, mean butt-log height (stem H), and mean tree height (H) for all walnut trees both in 2006 and 2009 divided in age classes Table 4.3: Walnut butt-log assortments obtained in 2009 with Nosenzo’s and Canesin’s classification in the 15 sample plots having walnut as principal species Table 4.4: Walnut butt-log assortments obtained in 2006 and in 2009 with Canesin’s classification in the 15 sample plots having walnut as principal species Table 4.5: Mean DBH, mean butt-log height (stem H), and mean tree height (H) for all wild cherry trees both in 2006 and 2009 divided in age classes Table 4.6: Wild cherry butt-log assortments obtained with Nosenzo’s and Canesin’s classification in the 19 sample plots having wild cherry as principal species Table 4.7: Wild cherry butt-log assortments obtained in 2006 and 2009 with Canesin’s classification in the 19 sample plots having wild cherry as principal species Table 4.8: Mean diameter at breast height (DBH), mean butt-log height (stem), and mean tree height (H) for all European ash trees divided in age classes both in 2006 and 2009 Table 4.9: European ash butt-log assortments obtained with Nosenzo’s and Canesin’s classification in the 10 samples plots having European ash as principal species Table 4.10: European ash butt-log assortments obtained in 2006 and 2009 with Canesin’s classification in the 10 samples plots having European ash as principal species. 5.

(6) Acknowledgement I would like to express my gratitude to Professor Antonio Nosenzo who has provided me with essential suggestions about the implementation of the stem quality classification adopted in this study. My gratitude goes also to my family, my uncle and my grandmother for their unforgettable support.. 6.

(7) Summary Thanks to the financial incentives available through the EEC Regulation 2080/92, almost 250 hectares of tree-farming plantations have been established on previous agricultural lands in Gorizia province (northern-east Italy). The main reasons of such a policy were the excess of agricultural productions and the lack of local wood supply. The amount of subsidies was significant but the rules enforced were not sufficiently strict and properly defined; as a consequence, several management problems occurred already during the seasons after planting. The aim of this study is to assess the present quality and the quality development of the tree farming plantations that were established in Gorizia province thanks to the public funds available through the EEC Regulation 2080/92. This study has included field visits, collection of sample data representative of the entire population and primary data analysis. Two methods of stem quality assessment have been adopted, namely Nosenzo’s classification (Nosenzo et al. 2008) and Canesin’s classification (2006). Both methodologies demonstrated to be quickly applied in the field but a statistical test was carried out to compare their outcomes. The results show the significant difference between Nosenzo’s and Canesin’s classifications. In particular, the second one considerably overestimates the amount of timber belonging to the higher quality classes (A and B classes) because of the less restrictive and more subjective parameters used. On the other hand, Nosenzo’s classification takes into account different degrees of some stem features (like 4 different classes for stem straightness); therefore, Nosenzo’s methodology demonstrated to be more unbiased and should be preferred. However, the results obtained with Canesin’s classification were used to investigate the stem quality development through the comparison of the 2009 results with the corresponding data obtained in the same sample plots in a previous study (which adopted the same classification methodology). The results suggest that there was a considerable decrease of walnut stem quality, because of the huge drop of the higher stem quality classes. Dissimilar situation occurs in both cherry and ash populations; a significant difference of the 2006 and 2009 butt-log grade distributions was found also there, but the amount of the best stem quality class has remained constant, while the. 7.

(8) others have changed; in particular, the reduction of the second best quality class (B class) has been compensated by the drop of the lowest class as well; therefore, the overall cherry and ash stem quality has decreased only slightly. Furthermore, the comparison between the 2006 and 2009 stand parameters demonstrates a general decrease of tree growth increments in the relatively oldest plantations; this is mainly due to the increasing competitive conditions among trees, the unsuitable species planted, and, in some cases, the poor soil condition on which they were established. On the other hand, in the youngest plantations the DBH current annual increment (1-2 cm/year) is usually higher than the mean annual increment, meaning that the negative competition among trees has not been reached yet. Lastly, tree-farming management in Gorizia province resulted to be scarce and inappropriate. Even in the thinned plantations the stem quality distribution has not improved due to the late and incorrect pruning operations that caused several butt-logs to fall in lower quality classes. Only in one-third of all plantations, the most valuable butt-logs represent more than 20 % of the total; this is the hypothetical value under which the profitability of a tree-farming plantation becomes questionable. This negative result is indicator of a general low interest in the active management of the tree-farming plantations established with the EEC Regulation 2080/92 and of low technical competences of the tree-farmers in Gorizia province.. 8.

(9) 1.. 1.1. Introduction. Background. The EEC Regulation 2080/92 was adopted to promote both sustainable farming and good silvicoltural practices. In those years, most EU countries (with the exception of France) registered a low local timber supply for their wood industries and high imports of commercial logs coming from tropical countries. As a result, the Regulation 2080/92 gave great impetus to the withdrawal of agricultural lands from food production in favour of trees able to produce high quality timber; in fact, planting trees on agricultural lands was one of the most important measures implemented. The aims were the reduction of current agricultural surplus and the increase of local timber availability. Italian farmers have hardly planted trees on their lands for productive purposes, with the only exception of the highly mechanized poplar plantations in northern-Italy. The large amount of money suddenly available pushed some landowners to accept the presence of trees on their lands; unfortunately, only few farmers had the appropriate knowledge and capacity to undertake such a new activity and also research on tree-farming plantation was at its early stages. Moreover, the wood industry were in favour of only certain timbers and the market influenced the choice of tree species used; this is why most plantation designs considered only European walnut and wild cherry as principal species, even on unsuitable soil conditions.. 1.2. Problem statement. Most of the tree farming plantations that have been established with the EEC Regulation 2080/92 have sometimes been managed in inappropriate ways or not managed at all. The amount of subsidies was significant and included planting costs, seedlings establishment costs (for 5 years after planting), and a premium covering losses of income (for 20 years after planting) originated from the crop that was previously cultivated on the same land; unfortunately, the management rules to accomplish were not sufficiently strict and properly defined and the farmers that became suddenly “tree growers” had not the appropriate knowledge and experience to manage properly their 9.

(10) new plantations. In some cases even abandonment or quasi-abandonment occurred; as a consequence, several management problems occurred already during the seasons after planting mainly due to the inappropriate set of tree species used, inappropriate plantation design, and lack of correct treatments like pruning, weed control, and thinning operations to be carried out at the right time. The trees planted grew up but only in a few cases their trunks are capable of producing high quality timber for veneers or high quality sawn-wood (for example, timber for high quality furniture). Grading the stem quality since the first stages is a proper way to monitor the quality development of a tree-farming plantation. Usually, commercial timber is graded visually by the personal experience of loggers and timber buyers, but estimating the quality of standing trunks is not always straightforward and a scientific methodology is needed, also to allow for unbiased comparisons.. 1.3. Objectives and research questions. In this study two different stem quality classifications were adopted: 1. the classification delineated by Nosenzo et al. (2008) based on measurable stem features and different classes for some relevant defects (for example, 4 classes of stem straightness and 3 classes related to the presence of branches and knots); 2. the classification used in a previous study carried out to assess the structure and the quality of almost all tree-farming plantations established in Gorizia province with the EEC Regulation 2080/92 (Canesin 2006, Canesin & Pividori 2007a and 2007b). Therefore, the specific research objectives are: • comparing and testing the differences between the two stem quality classifications adopted, namely Nosenzo’s and Canesin’s stem quality classifications; • estimating the present butt-log grade distribution and the stem quality development in a sample of all tree farming plantations established in Gorizia province with the EEC Regulation 2080/92.. 10.

(11) 1.4. Structure of the thesis. Chapter 1 introduces the reader with the background information that has motivated the present study. The research problems and objectives are stated. Chapter 2 highlights the relevant theories behind the study including some relevant knowledge about tree-farming plantations in Italy, most used tree species and plantation designs, and most important management operations that are supposed to be carried out in order to get high quality timber at the end of the productive cycle. Chapter 3 describes the research methodology including the description of the research approach, the description of the study area and sample plots, the instruments used for measuring the quantitative data, the data collection procedure and the final data analysis applied. Chapter 4 reports the detailed results of the study divided according to the three most used principal species: European walnut, wild cherry, and European ash. In each subchapter, the results obtained are also discussed. The detailed stand parameters and butt-log grade distribution obtained in 2009 for each sample plot are reported in Annex 4. Finally, the general conclusions that have been extrapolated from the results of this study are reported in Chapter 5.. 11.

(12) 2.. Theoretical background. The expression “tree-farming plantation” refers to the activity of growing trees with the purpose of producing commercial timber in a profitable way. These plantations are established temporally on agricultural lands and after the harvesting operations the land can be converted again into cropland without any restriction (Buresti & Mori 2003). The Public administration fostered the implementation of tree-farming plantations to accomplish the following objectives: • Reduce agricultural surplus; • Increase local wood production, since EU countries and above of all Italy are heavy importers; • Foster people to work and live also in marginal areas; • Reduce the amount of CO2 in the atmosphere, which is considered one of the main causes of the ongoing climate change; • Improve the environment quality thanks to the positive effects that a population of trees intrinsically bears (phytodepuration, landscape diversification, increased biodiversity, constitution of acoustic green barriers, recovery of rare ecological niches for wildlife, new jobs opportunities) (Buresti & Mori 2000). Also from the farmer point of view there are specific reasons that can lead toward the implementation of a tree-farming plantation: • Availability of public subsidies; • Diversification of both production and risks related to monocultures; • Utilization of agricultural lands that have become marginal and not longer suitable for conventional agricultural crops; • Production of renewable energy, like firewood and woodchips obtained after thinning that can be sold or used directly in the farm; • Production of non-timber-forest-products like honey and small fruits. Moreover, in the medium-long term it is forecasted a reduction of the incentives for agricultural crops and at the same time an increase of wood demand and wood prices (Buresti & Mori 2000). The principal factors that can determine the success or the failure of a tree-farming plantation are: the productive objectives of the farmer/land owner, the ecological features, and the socio-economic environment in which the plantation will be 12.

(13) established. Moreover, it is very important to entrust the project to a qualified treefarming plantation designer; in fact, the investments protract over a long time span (1015 years) and any mistake at the early stages should be avoided to maximize the production at the end of the rotation (20-40 years). Therefore, before planting operations, the following aspects and activities should be considered and carefully evaluated: • Data collection regarding owner perspectives and farm features; • Analysis of the soil characteristics of the land that is going to be planted, including soil type, spontaneous flora, and exposure; • Definition of a suitable plantation design; • Evaluation of the need of accessories aimed to protect the seedling or to foster their growth, like dark plastic films, shelters, supporting poles, and fencing; • Providing the necessary documentation to be presented to local authorities; • Getting the project approval. After the approval of a well described project, it is possible to proceed with the practical realization that include the main following steps: • Ordering the seedlings (better with local provenances); • Carrying out hydraulic works when needed; • Carrying out soil preparation during summer time when the soil has the appropriate physical properties (neither wet nor very dry); • “Squaring” the land in order to distribute at the specified distance the seedlings; • Planting the seedlings at the right period according to climate and soil conditions (Buresti & Mori 2000). The most common tree species that can produce high quality timber and get up to high prices in the European market are: wild cherry (Prunus avium L.), pear (Pyrus spp.), Sorbus spp., maple-trees (Acer spp.), chestnut (Castanea sativa Miller), European ash (Fraxinus excelsior L.) and European walnut (Juglans regia L.), which produces the most requested commercial timber in the Italian market (Buresti & Mori 2004). However, the prices depend not only on the wood species but also on the wood mechanical quality; depending on the amount and type of defects found on a commercial log, the timber price changes accordingly and different quality classes can be delineated. Sometimes, the price for high quality timber is not divided into a specified number of quality classes but it is given as range; this is the case of “Camera. 13.

(14) di Commercio Udine” (2009), that reports the price of walnut timber in a range going from a minimum of 413 euro/m3 to a maximum of 826 euro/m3. Diversely, according to Buresti & Mori (2004), walnut industrial logs may be classified in three price classes as reported in Table 2.1. Table 2.1: Prices of walnut logs belonging to different grades in 2003 (Buresti & Mori 2004) Euro/m3 1st class (A). 1100. 2nd class (B). 350. rd. 3 class (C). 160. To be classified as first class, logs should be at least 30-40 cm in diameter and 2.5 m long; they should have a straight and cylindrical shape, all knots and defects included in the 10 cm diameter central cylinder, homogeneous wood colour and regular annualgrowth rings. The top quality timber (A class) can be used in the veneer industry (both sliced veneer and rotary-cut veneer). Depending on the presence and amount of the defects that affect the ideal features mentioned above, a log is downgraded to the second or third class (Buresti & Mori 2004). Second class timber (B class) represents high quality sawn-wood, suitable for the production of high quality furniture, interior joinery, doors, turnery, etc. Third class timber (C class) represents low quality sawn-wood used only for exterior joinery, light construction, boxes, crates, etc. Finally, a log is classified as D class when it does not fit into class C either, due to its bad mechanical defects; class D timber represents the lowest quality grade and only firewood (logs cut into peaces) or woodchips that are burned for energy purposes can be produced out of it. In any tree-farming plantation, woody biomass for energy purposes is also a by-product when produced from branches of principal trees, and from stems and branches of shrubs and accessory trees (Figure 2.1; Buresti & Mori 2006).. Figure 2.1: Timber assortments that can be obtained from tree-farming plantations (Buresti & Mori 2000, modif.).. 14.

(15) 2.1. Definitions. Several technical expressions are used to describe properly tree-farming plantations. First of all, it is important to clarify the difference between “principal tree species” and “accessory tree species”. Principal trees are the “crop trees” in a plantation, i.e. those trees that are planned to produce the high quality timber at the end of the productive cycle; their trunk can be sold at the highest prices and the final profit depends mostly on them. However, it is possible and strongly suggested by recent researches (Buresti & Mori 2007a) to plant also accessory tree species; in fact, it has been demonstrated that they favour the formation of a suitable stem and crown architecture of the principal trees thanks to the shading effect, and the improvement of soil fertility; in addition, with thinning operations they produce woody biomass that can be sold as firewood or woodchips representing an additional income for farmers; shrubs, shade tolerant species and all those species able to improve soil fertility through nitrogen fixation and leaf litter are preferred. Both principal species and accessory species are planted following a precise plantation design, which is defined as the minimum unit of land surface including all tree species used and their relationship defined in the project design (like distance between trees of the same species and distance between trees of different species). The plantation design is able to reproduce on the entire land surface the whole plantation just rotating repeatedly itself by 180° on each side, without changing the relationship between species. The boundaries of each plantation unit pass through the centre of the trees located on the plantation design borders. Practically, thanks to the plantation design unit it is possible to know the amount of species needed, the number of trees of each species, the spacing distance and the spatial distribution of each tree species (Buresti & Mori 2000). The easiest plantation design (pure plantation) uses only one tree species planted at a fixed spacing. When there are two or more principal species the tree population is defined as mixed plantation. A plantation with only one principal species and one or more accessory species is denominated as pure plantation with accessory species, whereas, if there are two or more principal species plus the accessories species, the plantation is defined as a mixed plantation with accessory species (Figure 2.2 & Figure 2.3) (Buresti & Mori 2007b).. 15.

(16) Figure 2.2: Example of a mixed plantation with accessory species: plantation design (left) and same plantation after removal of all accessory species (right) (Buresti & Mori 2000, modif.).. Figure 2.3: Types of tree farming plantations according to species composition. Starting from left: pure plantation (plot n° 7, see Annex 4), mixed plantation (plot n° 30), pure plantation with accessory species (plot n° 53), and mixed plantation with accessory species (plot n° 16.1).. 2.2. Theoretical approaches. In order to get positive results at the end of the rotation, both qualification and dimensioning of all principal species should be carefully followed during the years. According to Buresti & Mori (2000), the three phases at the base of high quality timber production are: • Forming a well developed root system, that can be ensured through soil preparation, right selection of tree species, good quality seedlings, right plantation technique and suitable treatments during the first years after planting (emergency watering, weed control, replacement of dead seedlings); • Forming a trunk sufficiently long (at least 2.5 m), cylindrical and free of knots thanks to the right tree species selection, a periodical monitoring of each single principal tree and correct pruning operations when needed; • Fostering the formation of a cylindrical stem with homogeneous annual-growth rings in cross section, meaning a constant increment in size year by year (monitoring and thinning operation should be carried out at the proper time).. 16.

(17) The utilization of accessory species together with a mix of principal tree species has been proved to have several positive effects on the plantation management: • Faster soil cover and reduced soil erosion; • Improvement of soil fertility and biodiversity; • Positive modification of principal trees stem shape and architecture; • Better tree selection thanks to planned thinning operations; • Intermediate incomes derived from non timber forest products (honey, small fruits), firewood and small size sawn-wood obtained with thinning operations; • Reduction of both treatment inputs and management costs: less use of fertilizers, easier weed control (especially when shrubs are planted), easier pruning operations carried out only on a restricted number of trees (only on principal trees and not on the accessory trees); • Reduced pathogens’ attacks; • Diversification of production in terms of timber assortments and a consequent reduction of hazards due to market price fluctuations (Buresti & Mori 2000, Buresti & Mori 2003, Tani et al. 2007, Cutini & Giannini 2007). To ensure the formation of suitable stems for high quality timber, pruning plays a very important role. Its main purpose is to guide the correct formation of a straight stem free of knots and branches and to concentrate all the defects in the ideal central cylinder of 8-10 cm diameter (Buresti et al. 2007, Brunetti & Nocetti 2007) (Figure 2.4).. Figure 2.4: Ideal central cylinder containing all defects (Buresti & Mori 2000, modif.).. Thinning operations are also very important to be carried out at the right time when competition at the expenses of principal trees occurs (Pelleri et al. 2007; Marchino & Ravagni 2007). If thinning is delayed or not carried out, the entire plantation could be badly affected. In fact, all trees start to become weaker due to the increasing competition for nutrients and lights; this situation could also facilitate pathogens’ 17.

(18) attacks (fungi, insects and bacteria). Moreover there will be a delay in the productive cycle with respect to the original project design and the timber obtained could not be of the highest classes as planned (inhomogeneous annual growth rings). Thinning operations are still very expensive in young plantations because of the low amount of wood obtained and the low firewood quality that mostly comes from fast growing accessory species (like willow and alders) or accessory shrubs (like hazel). Research is trying to find a better market allocation for the timber harvested with thinning operations and some positive results have already emerged like: the profitable chairs production with plywood derived from small size walnut logs (Zanuttini et al. 2009) and the potential production of window frames with alder timber (Todaro et al. 2007). To estimate the value of a tree-farming plantation is a hard job and currently there are few professionals able to do it properly. It is very important to have knowledge about wood in general, the growing patterns of each tree species and the wood technological characteristics demanded by the wood industry. At the moment of final harvesting different timber assortments will be available and to maximize profits it is crucial to allocate each log to the most valuable quality class even though this could cause a reduction in the amount of commercial timber sold. For example, if there is a third grade log four meters long but with the first 2.5 metres of A class, it is much profitable to cut it down to 2.5 metres and sell it as A class (higher price) instead of C class (much lower price). As a conclusion, appropriate plantation design, pruning and thinning operations are key factors to get high quality timber at the end of the technical rotation (Buresti & Mori 2004).. 18.

(19) 3.. Materials and methods. 3.1. Research approach. In order to examine the quality development of broad-leaved tree-farming plantations in northern-east Italy, both secondary sources and primary data have been collected. The literature review allowed for a comprehensive understanding of the terminology used in the high quality tree-farming sector. Secondary data include the 2006 field data and results carried out to investigate the structure of all tree-farming plantations established in Gorizia province with the EEC Regulation 2080/92 (Canesin, 2006; Canesin & Pividori 2007a and 2007b). The balk of the study is based on primary data that have been collected in spring 2009 to test the differences between the two stem classifications adopted; both quantitative and qualitative data collected have been subjected to statistical analysis to investigate the growth and quality trend of tree-farming plantations in Gorizia province. To make reliable the comparison between the secondary data obtained in 2006 (Canesin, 2006) and the primary data obtained with the fieldwork carried out in 2009, the permanent plot approach was adopted.. 3.2. Study area. Gorizia province extends over 466 km2 and it is the most populated province in FriuliVenezia Giulia region with about 296 inhabitants/km2. It borders Slovenia to the east, Udine province to the west and the Adriatic see and Trieste province to the south (Figure 3.1). The territory can be divided in five landscape units: 1) Collio, which is located to the north on low mountains with a dense river network; 2) high plane, which has been formed by coarse sediments transported by the rivers; therefore, the soil has low retention capacity; 3) low plane, located south of the “resurgences” and has been formed by fine and very fine sediments (clay and silt); both soil fertility and water availability reach high levels; 19.

(20) 4) Grado lagoon, which is located to the south near the mouth of the Isonzo and Tagliamento rivers. Its origin is due to the redistribution of the alluvial sediments by the see. The salty water and the low nutrients in the soil are the main limiting factors for a well-developed vegetation cover. Most of the original salty and wet soils have been drained in the past to get more agricultural land available; the most significant example is the “Victoria drainage”, completed at the end of the 50’. 5) Carso, which represents a unique landscape unit characterized by pure calcareous rocks able to form particular morphologies like sinkholes and dolinas. The high vegetation is poor due to the high pressure of the army during the first world war and the lack of conifers plantations that were common in others areas after the wars (AA.VV. 1998; Abramo & Michelutti 1998).. Figure 3.1: Map of Gorizia province (source: mondimedievali 2009, modif.). The climate is mild thanks to the influence of the Adriatic See. The annual rainfall is higher in the interior part reaching even 1600 mm, while it hardly reaches 1000 mm on the coastal area. Also the mean annual temperature changes between the interior and coastal zones going from 12°C to 14°C respectively. The “Carso” area, the low plane and the lagoon are particularly affected by a very strong wind blowing from northeast and called “Bora” (AA.VV. 1995).. 20.

(21) 3.3. Sample plots. In 2006, Canesin (2006) carried out a study on the structure of almost all broad-leaved tree-farming plantations realised in Gorizia province with the EEC Regulation 2080/92 (about 250 ha). From the 147 homogeneous areas visited in 2006, a sub-sample of 21 sample plots has been selected choosing from the relatively best plantations spread on the entire province; therefore, the sample plot identification numbers correspond to those used in Canesin’s study to allow for easier comparison; Table 3.1 lists the plots number and the municipalities in which they are located. Each sample plot was assumed to represent significantly either the whole plantation it is located in or only a homogenous area inside the plantation itself (in this case the plot number bears the homogeneous area number after the point). Most of the plantations are mixed with accessory species (14 plots); three are mixed, three pure with accessory species and only one is a pure plantation. Table 3.1: List of the 21 plots sampled in 2009 with the respective species composition and municipality in which they are located. Plot n° 1 2.1 2.2 7 12A 15A 16.1 16.2 22 30 35 37A 41.1 41.2 51.1 51.2 53 54A 61 77A.1 77A.2. Municipality Romans d'Isonzo Cassegliano Cassegliano Grado San Canzian d'Isonzo San Lorenzo Isontino San Pier d'Isonzo San Pier d'Isonzo Turriaco Mossa San Canzian d'Isonzo San Canzian d'Isonzo Grado Grado Romans d’Isonzo Romans d’Isonzo Grado Dolegna Grado Romans d’Isonzo Romans d’Isonzo. Composition Mixed Mixed with accessory species Mixed with accessory species Pure Mixed with accessory species Mixed with accessory species Mixed with accessory species Mixed with accessory species Mixed Mixed Mixed with accessory species Mixed with accessory species Pure with accessory species Pure with accessory species Mixed with accessory species Mixed with accessory species Pure with accessory species Mixed with accessory species Mixed with accessory species Mixed with accessory species Mixed with accessory species. 21.

(22) Reading from Canesin’s field data, it was possible to localize the trees already measured in 2006 through the combination of the recorded row number in the plantation and the assigned tree number along the row. Only in three cases it was not possible to find exactly the same tree measured due to missing data in the 2006 field papers (this applies to plots n° 7, 15A and 16.2). The plot shapes and sizes are variable: 30 trees per each principal species have been measured in 2006 along the most representative rows of each plantation. These 30 trees per each principal species have been localized and measured in 2009 as well. Slight differences occur in those plantations where a recent thinning operation has been carried out (this applies to plot n° 2.1, 2.2, 16.1, and 16.2). In this case some trees in the nearest rows were sampled to reach totally 30 trees per each principal species in each plot.. 3.3.1 Instruments for measuring Both spacing and plantation designs were drawn measuring with a metric tape the minimum distance between trees and principal trees in each plot. Then, for each sampled tree, a common tailor’s tape was used to measure the girth at breast height, and a metric pole to measure the butt-log height and the deviation from the straightness. The tree total height was estimated using the Vertex hypsometer. Finally, the trunk quality features have been recorded following systematically the specific guidelines of both classification methodologies adopted.. 3.3.2 Data collection For each sampled tree the girth at breast height, the butt-log height, the total height and the trunk deviation from straightness were measured and the stem quality features were described filling in the field paper reported in Annex 2. The deviation from the straightness was divided in 4 classes according to Nosenzo’s classification (Nosenzo et al. 2008): class 0 means that there is no deviation (the buttlog is perfectly straight), class 1 means a deviation between 1 and 3 %, class 2 a deviation between 3 and 10 % and class 3 a deviation more than 10 %. The stem inclination was divided into 2 classes: class 1 means a stem inclination between 10 and 20 % while class 2 means an inclination more than 20 %.. 22.

(23) Figure 3.2: Stem straightness classes in Nosenzo’s classification (Nosenzo et al. 2009, modif.). It was also recorded, per each butt-log sampled, the presence of branches and knots with diameter more than 3 cm, covered knots, fresh knots and rotten knots. The sum of the diameters of knots and branches was divided in 3 classes according to Nosenzo’s classification: class 1 when the sum of all knots and branches diameters is more than 60 mm, class 0.5 when the sum is between 15 and 60 mm and class 0 when less than 15 mm. Protuberances caused by stubs were considered as knots with a diameter of 60 mm due to their negative effect on wood quality.. Figure 3.3: Some stem defects considered in this study. From left to right: biotic defects (insect holes, cancer and rot); a-biotic defects (knot diameter > 3 cm and base damage). The a-biotic defects in these examples could not be considered negligible because the stem diameter (at the height where the defect is found) was more than 10 cm.. Moreover, for each stem it was recorded the presence of sinuosity, fork, “saddle”, base damage, rots, insects holes, frost cracks, ovality, signs on bark, “bottle neck”, spiral. 23.

(24) grain and others specific biotic or a-biotic defects reported in the notes (like dominancy, debarking, superficial insect holes, metal wires in the wood, etc.). Some of the above defects were not found in the English literature and a word-by-word translation was applied; to avoid misunderstandings, the Italian translation and a short description of each stem defect treated in this study is reported in Annex 1.. 3.4. Data analysis. The 2009 field data were analyzed using the Microsoft Office Excel software. At first, the raw field data of each sample plot have been recorded in an Excel sheet; then, a new excel sheet was created for each principal tree species found in the plot and the respective descriptive statistics have been computed. To grade the butt-log of each sampled tree, two different methods were used: Nosenzo’s classification (Nosenzo et al. 2008) and Canesin’s classification. Both methods above divide the stem quality in 4 classes: • A class, which is the most valuable one; the butt-logs belonging to this class have high quality features and can be used in the veneer industry; • B class, which includes valuable timber that can be used for high quality furniture, for instance; • C class, which includes butt-logs with low quality features and therefore capable of producing only low quality sawn-wood; • D class, which is the lowest quality; the trunks belonging to this class can be used only as firewood or bioenergy (woodchips). In both classifications adopted, a stem falls in D class whenever it is shorter than 2.5 meters, or it has got a deviation from the straightness above 10%, or its wood is affected by either serious biotic defects (insect holes, rot, serious debarking, etc.), or severe abiotic defects (significant mechanical damages at the stem base able to facilitate disease attack). Superficial insect holes affecting only the bark and not the wood were considered negligible (Figure 3.4). Additionally, D class includes also stems of trees that show sinuosity or an inclination above 20 % in Nosenzo’s classification, while, according to Canesin’s classification, a stem is considered D class also when it has got both deviation and sinuosity. Whenever all the conditions above do not apply, a stem can be included in a class other than D. 24.

(25) Figure 3.4. Tiny insect holes that affect only the bark of some European ash stems. This defect has been considered negligible.. According to Nosenzo’s classification, a stem belongs to C class when its deviation from the straightness is between 3 and 10 %, or the sum of the diameters of all branches and knots on the butt-log is more than 60 mm. Instead, according to Canesin’s classification, a stem is considered C whenever it has got deviation or sinuosity (not both together), or both branches and knots with a diameter larger than 3 cm. Table 3.2: Comparison between the 2 classifications adopted in this study. The occurrence of at least one feature listed in column 2 and 3 determines the corresponding grade in column 1. Class. Nosenzo’s classification. Canesin’s classification. D. Stem length < 2.5 m. Stem length < 2.5 m. D. Deviation from the straightness > 10%. Both stem deviation AND sinuosity. D. Relevant biotic or a-biotic defects. Relevant biotic or a-biotic defects. D. Stem inclination > 20%. -. D. Stem sinuosity. -. C. Deviation from the straightness in the. Stem deviation OR sinuosity. range of 3-10% C. B. Sum of the diameters of all branches and. Presence of knots AND branches with a. knots > 60 mm. diameter > 3 cm. Deviation from the straightness in the. Absence of stem deviation AND sinuosity. range of 1-2% Sum of the diameters of all branches and. Presence of knots OR branches with a. knots in the range of 15-60 mm. diameter > 3 cm. A. Deviation from the straightness < 1%. Absence of deviation AND sinuosity. A. Sum of the diameters of all branches and. Absence of knots AND branches with a. knots < 15 mm. diameter > 3 cm. B. 25.

(26) If the deviation from the straightness is between 1 and 3 %, and the sum of the diameters of all branches and knots on the butt-log is between 15 and 60 mm, the stem falls in B class in Nosenzo’s classification. Instead, according to Canesin’s, B class is reserved for butt-logs that show neither deviation, nor sinuosity, nor a contemporaneous presence of both branches and knots having a diameter more than 3 cm. Finally, a stem is considered A class only if its deviation from the straightness is less than 1% and the sum of the diameters of all branches and knots on the trunk is less than 15 mm (Nosenzo’s). Instead, according to Canesin’s, it falls in A class whenever it is straight and free of both branches and knots with a diameter larger than 3 cm. Table 3.2 lists the above conditions to be fulfilled in both classification methodologies. The main implementation problem was the relatively subjective interpretation of the “stem deviation” in Canesin’s classification. Afterwards, pie charts showing both the butt-log grade distribution in each plot for each species and the overall butt-log grade distribution for each species have been drawn. The statistical Fisher exact test was carried out to investigate both the similarity between the two stem classifications used and the development of the butt-logs quality in the period 2006-2009. In the first case, per each sample plot the number of trunks belonging to each quality class obtained through Nosenzo’s classification was compared with the corresponding number of trunks obtained with Canesin’s classification. In the second case, per each sample plot the number of trunks belonging to each quality class obtained through Canesin’s classification in 2006 was compared with the number of trunks belonging to the same quality class obtained with the same classification in 2009. The corresponding probability values were computed using the online software provided by “College of Saint Benedict and Saint John’s university” (Kirkman 1996). The Fisher exact test was required due to the high frequency of expected values less than 5, which is the minimum expected value found in the literature that allows for the accurate application of the Chi-square test (Fowler & Cohen, 2002). However, the Chisquare test was applied to investigate the statistical difference between the 2006 and 2009 results concerning the number of all butt-logs falling in a certain quality class and belonging to the same principal species (in this case, the expected values were much higher than 5 due to the sum of all butt-logs of a certain quality class in each plot); STATISTICA software was used to compute the Chi-square value and the corresponding probability of occurrence.. 26.

(27) 4.. Results and Discussions. The fieldwork carried out in spring 2009 produced a consistent amount of raw data. A first analysis aimed at the computation of the stand parameters for each sample plot of this study (21); the detailed results are reported in Annex 4, where the reader can find information about the year of plantation, the species composition, and the specific plantation design. Per each principal species there is also a table reporting the growth increment, a table with the stand parameters and a table with the percentage of all stem defects found. Finally, the butt-log grade distribution obtained with both Nosenzo’s and Canesin’s stem classifications are presented in pie charts for ease of comparisons. The most represented principal tree species in Gorizia province are European walnut, wild cherry, and European ash; the results concerning each principal tree species are described and discussed separately in 3 different sub-chapters focusing on: 1. the stand parameters estimated in 2009 and the current annual increment referred to the last three years; 2. the most relevant stem defects found; 3. the share of each stem quality classes obtained with Nosenzo’s and Canesin’s classifications and their statistical comparison through the Fisher exact test; 4. the stem quality development occurred between the years 2006 and 2009. The percentage of each single stem defect referred to the total sample size of each principal tree species is shown for ease of comparison in Table 4.1; taking as example the first value, the table reports that 9 % of all walnut trees show at least one branch with a diameter larger than 3 cm. As one single butt-log could present even all defects listed, the values in Table 4.1 cannot be summed up inside the same tree species group. In all three principal species, signs on bark and covered knots are the most common defects followed by stem straightness between 3 and 10 % and sum of all branches and knots diameters more than 60 mm ( (b+k) > 60 mm). All defects except signs on bark (because considered a negligible defect) were divided into 5 groups according to the origin and the stem feature they affect (see Annex 1 for a detailed stem defects description).. 27.

(28) Table 4.1: Percentage of each stem defect found on all sampled trees of each principal species and grouped according to the origin and the stem feature they affect. Stem defect. Braches and knots. Stem shape. Biotic A-biotic Grain Negligible. Branch > 3 cm Knot > 3 cm Covered knots Fresh knots Rotten knots (b+k) > 60 mm (b+k) 15mm < > 60 mm Stem straightness > 10% Stem straightness 3% < > 10% Stem straightness 1% < > 2% Stem inclination > 20% Stem inclination 10% < > 20% Sinuosity Fork Saddle Bottle neck Ovality Rot Insect hole Base damage Frost crack Spiral grain Fluting Signs on bark Sample size (n°). 4.1. Walnut (%) 9 31 64 32 2 57 3 11 61 26 5 28 5 3 21 2 0 0 5 15 2 0 1 93. Cherry (%) 13 12 59 17 2 45 11 3 29 52 1 7 1 4 12 4 1 2 0 16 2 1 0.4 88. Ash (%) 14 3 39 5 1 30 7 3 31 50 0 4 2 2 21 0 0 0 8 11 1 0 0 90. 447. 570. 300. European walnut (Juglans regia L.). All sample plots having walnut as principal species were divided in age classes and the corresponding mean tree height annual increment and mean diameter at breast height (DBH) annual increment were computed. Because of the very diverse management and ecological situations found (different plantation design, different soil conditions, different seedlings quality, different treatments) the expected DBH increase with age does not occur; in fact, one of the lowest mean DBH (10.7 cm) was found in the oldest plantations. However, the highest mean DBH (16.1 cm) was found in plantation number 7, which is the only pure walnut plantation of this study. The highest mean tree height was found in plot number 51.1 (12.31 m) instead, which is a mixed plantation with accessory species three years younger than the previous pure plantation. 28.

(29) The mean DBH annual increment is in the range of 0.7-1.4 cm/year with the highest value found in the 9 years old plots, while the tree height annual increment goes from 0.59 m/year in the oldest plots to 1.18 m/year in the 9 years old plots. Afterwards, the current annual increment (CAI) referred to the last three years was computed: values lower than the mean annual increment are found mostly in the older plantations meaning that a negative competition occurred and the trees produced thinner annual growth rings during the last years. Same trend for the height CAI, which is higher than the mean height annual increment only in the 8 years old plantations (the youngest). The detailed data are reported in Table 4.2. Table 4.2: Mean DBH, mean butt-log height (stem H), and mean tree height (H) for all walnut trees both in 2006 and 2009 divided in age classes. The last four columns report the mean DBH and height annual increment (Im) and the mean DBH and height current annual increment (CAI, period 2006-2009).. 27 30 30 87. 2006 DBH H (cm) (m) 9.3 6.38 9.1 7.15 7.9 6.53 8.8 6.69. DBH (cm) 13.2 10.5 8.6 10.7. 2009 Stem H (m) 2.53 2.65 2.76 2.65. H (m) 9.17 8.04 6.76 8.28. dbh (cm/y) 0.94 0.75 0.61 0.76. h (m/y) 0.66 0.57 0.48 0.59. Im. CAI dbh h (cm/y) (m/y) 1.29 0.93 0.46 0.30 0.25 0.08 0.64 0.53. 1 2.1 2.2 Mean. Age (y) 14 14 14 14. 7 12A Mean. 12 12 12. 30 30 60. 11.8 11.1 11.5. 7.18 7.90 7.54. 16.1 14.9 15.5. 3.00 2.61 2.81. 8.33 10.31 9.49. 1.34 1.24 1.29. 0.69 0.86 0.79. 1.43 1.28 1.35. 0.38 0.8 0.65. 16.1 16.2 Mean. 11 11 11. 30 30 60. 10 8.4 9.2. 6.57 5.87 6.22. 11.9 11.6 11.8. 2.37 2.37 2.37. 6.85 7.24 6.99. 1.08 1.05 1.07. 0.62 0.66 0.64. 0.65 1.07 0.86. 0.09 0.46 0.26. 22 30 Mean. 10 10 10. 30 30 60. 8.8 6.3 7.6. 7.32 4.05 5.7. 12.1 11.7 11.9. 3.59 1.99 2.79. 8.39 6.67 7.51. 1.21 1.17 1.19. 0.84 0.67 0.75. 1.11 1.78 1.44. 0.36 0.87 0.61. 35 37A 51.1 51.2 Mean. 9 9 9 9 9. 30 30 30 30 120. 8.6 10.8 8.2 9.2. 6.42 9.42 6.83 7.56. 12.3 14.6 12.7 11.1 12.7. 3.61 2.67 2.50 2.32 2.78. 9.06 11.67 12.31 9.80 10.58. 1.36 1.62 1.41 1.24 1.41. 1.01 1.30 1.37 1.09 1.18. 1.24 1.27 1.51 1.34. 0.88 0.75 1.82 1.15. 77A.1 77A.2 Mean. 8 8 8. 30 30 60. 5.7 3.2 4.5. 4.92 3.20 4.1. 11.0 6.2 8.6. 2.36 2.21 2.29. 9.22 5.19 7.94. 1.38 0.78 1.08. 1.15 0.65 0.99. 1.76 1.01 1.38. 1.43 0.66 1.29. Plot n°. N°. Figure 4.1 plots the mean DBH annual increment of each plot against the age and shows how the mean annual increment decreases when the plantations become older. The polynomial trendline that was drawn illustrates the phenomena. To keep constant the. 29.

(30) mean annual increment, negative competition between trees should be controlled through appropriate thinning operations. Some of the oldest plantations (plots 2.1, 2.2, 16.1, and 16.2) were thinned in the winter 2007/2008 (1 year before this study): mostly accessory species were affected and only very few walnut trees were felled. Consequently, more light and nutrients are currently available and a slight increase of both mean annual increment and CAI is expected to occur in these four plantations during next years.. Figure 4.1: Mean DBH annual increment of walnut trees plotted against age.. The stem defects found in all walnut trees were divided into 5 groups as shown in Table 4.1. Signs on barks are very common (93%), but because they could be considered negligible at this stage, they are not included in further analysis. It was estimated that in average each walnut tree shows 4 different defects; this estimate should be carefully interpreted but suggests that the amount of defects found was quite huge. According to the results obtained, the most common stem defects are due to the presence of knots and branches and to the stem shape. In both cases a correct and constant pruning would have lowered very much the amount of this two defect groups. Biotic (insect holes and rot) and grain defects (spiral grain and fluting) share a much lower percentage of the total, while the a-biotic defects (base damage and frost crack) present a relatively low percentage but were found on 78 butt-logs (17% of the sample size); therefore, they cannot be considered irrelevant (Figure 4.2).. 30.

(31) Figure 4.2: Stem defects found on all walnut butt-logs. See Table 4.1 for the list of stem defects included in each group.. Figure 4.3 shows the contribution of each single defect inside the defect group “knots and branches”: the most common are covered knots followed by the sum of branches and knots diameters more than 60 mm, branches and knots with diameter larger than 3 cm and fresh knots. This demonstrates that pruning has not been carried out in the proper way and at the right time. Rotten knots share only 1 % of this defect group.. Figure 4.3: Stem defects due to the presence of knots and branches on all walnut butt-logs.. The detailed 2009 butt-log grade distribution in each sample plot having walnut as principal species are reported in Table 4.3, while the total percentage of A, B, C and D classes for the entire walnut population is reported in Figure 4.4. In all plots, Nosenzo’s classification produces lower percentage of higher quality classes (A and B) when compared with Canesin’s classification, but the differences. 31.

(32) between the two classifications are significant only in two cases (plot 22 and 35) when applying the Fisher Exact test. Table 4.3: Walnut butt-log assortments obtained in 2009 with Nosenzo’s and Canesin’s classification in the 15 sample plots having walnut as principal species. The last two columns show the p value and the equivalent significance level obtained with the Fisher exact test (*=little significant; **=significant; ***=very significant, n.s.=not significant) Nosenzo's classification. Canesin's classification. Plot n°. Age (y). A. B. C. D. A. B. C. D. p (Fisher). Significance. 1. 14. 0%. 11%. 33%. 56%. 7%. 0%. 44%. 48%. 0.166. n.s.. 2.1. 14. 0%. 0%. 37%. 63%. 3%. 3%. 33%. 60%. 1.000. n.s.. 2.2. 14. 0%. 7%. 63%. 30%. 7%. 10%. 50%. 33%. 0.523. n.s.. 7. 12. 0%. 0%. 43%. 57%. 3%. 3%. 40%. 53%. 1.000. n.s.. 12A. 12. 0%. 0%. 53%. 47%. 0%. 10%. 50%. 40%. 0.324. n.s.. 16.1. 11. 0%. 7%. 60%. 33%. 3%. 3%. 60%. 33%. 1.000. n.s.. 16.2. 11. 0%. 0%. 40%. 60%. 0%. 3%. 37%. 60%. 1.000. n.s.. 22. 10. 0%. 20%. 57%. 23%. 17%. 0%. 67%. 17%. 0.005. ***. 30. 10. 0%. 3%. 10%. 87%. 7%. 0%. 7%. 87%. 0.671. n.s.. 35. 9. 0%. 0%. 60%. 40%. 3%. 13%. 43%. 40%. 0.098. *. 37A. 9. 3%. 7%. 70%. 20%. 10%. 0%. 70%. 20%. 0.491. n.s.. 51.1. 9. 0%. 0%. 57%. 43%. 3%. 0%. 60%. 37%. 0.792. n.s.. 51.2. 9. 0%. 0%. 37%. 63%. 3%. 0%. 37%. 60%. 1.000. n.s.. 77A.1. 8. 0%. 3%. 37%. 60%. 0%. 0%. 40%. 60%. 1.000. n.s.. 77A.2. 8. 0%. 7%. 37%. 57%. 7%. 0%. 37%. 57%. 0.393. n.s.. Figure 4.4: Walnut butt-log grade distribution according to Nosenzo’s classification (left) and Canesin’s classification (right).. “A class” timber is almost 0% of the total according to Nosenzo’s classification and not more than 5 % according to Canesin’s; “B class” ranges between 3 and 4%, while almost half of all walnut trees fall in “D class” according to both classifications used. As a consequence, the amount of high quality timber that is still potentially suitable for veneers and high quality sawn-wood at the end of the rotation will be very little.. 32.

(33) Therefore, the economical benefit will be strongly compromised. In fact, only in one case (plot 22) the amount of A and B classes together reaches 20 %; this is the threshold value found in the literature (Nosenzo et al. 2008) under which the economical benefit is no longer met and the future management of the entire plantation becomes questionable. To investigate the stem quality development occurred in the period 2006-2009, the Fisher exact test was applied to the 2006 and 2009 quality classes obtained with Canesin’s classification (Table 4.4). Plot 51.2 was not considered in this analysis because of the lack of 2006 data. In 85 % of the cases (12 plots out of 14) it was found a significant difference between the 2006 and 2009 values and even a very significant difference in 64 % of the cases (9 plots out of 14). These results suggest that the stem quality has changed consistently during the last three years. Table 4.4: Walnut butt-log assortments obtained in 2006 and in 2009 with Canesin’s classification in the 15 sample plots having walnut as principal species. The last two columns show the p value and the equivalent significance level obtained with the Fisher exact test (*=little significant; **=significant; ***=very significant, n.s.=not significant) 2006 (Canesin’s class.). 2009 (Canesin's class.). Plot n°. Age (y). A. B. C. D. A. B. C. D. p (Fisher). Significance. 1. 14. 10%. 27%. 7%. 57%. 7%. 0%. 44%. 48%. 0.000. ***. 2.1. 14. 13%. 0%. 20%. 67%. 3%. 3%. 33%. 60%. 0.256. n.s.. 2.2. 14. 27%. 7%. 23%. 43%. 7%. 10%. 50%. 33%. 0.066. *. 7. 12. 10%. 13%. 7%. 70%. 3%. 3%. 40%. 53%. 0.007. ***. 12A. 12. 30%. 13%. 17%. 40%. 0%. 10%. 50%. 40%. 0.001. ***. 16.1. 11. 20%. 7%. 10%. 63%. 3%. 3%. 60%. 33%. 0.000. ***. 16.2. 11. 10%. 0%. 3%. 87%. 0%. 3%. 37%. 60%. 0.000. ***. 22. 10. 33%. 27%. 17%. 23%. 17%. 0%. 67%. 17%. 0.000. ***. 30. 10. 33%. 0%. 3%. 63%. 7%. 0%. 7%. 87%. 0.028. **. 35. 9. 7%. 20%. 27%. 47%. 3%. 13%. 43%. 40%. 0.617. n.s.. 37A. 9. 17%. 23%. 10%. 50%. 10%. 0%. 70%. 20%. 0.000. ***. 51.1. 9. 3%. 3%. 3%. 90%. 3%. 0%. 60%. 37%. 0.000. ***. 51.2. 9. -. -. -. -. 3%. 0%. 37%. 60%. -. -. 77A.1. 8. 3%. 0%. 13%. 83%. 0%. 0%. 40%. 60%. 0.039. **. 77A.2. 8. 3%. 0%. 7%. 90%. 7%. 0%. 37%. 57%. 0.006. ***. A further analysis of the total amount of all butt-logs falling in the four quality classes in both 2006 and 2009 shows that there was a consistent decrease of the higher quality classes in favour of C class. 33.

(34) The Chi-square test was applied and a very significant difference between the 2006 and the 2009 butt-log grade distribution emerged (X2 = 130; p value = 0.000). In fact, “A class” has decreased consistently from 16 to 5 % and “B class” from 10 to 3 %. The main causes are due to the presence of branches and knots with a diameter larger than 3 cm. On the other hand, the reduction of “D class” butt-logs is probably due to the treesgrowth occurred during the very dynamic stem formation stage: for example, it could happen that butt-logs shorter than 2.5 m were considered D class in 2006 but they then developed in longer stems during the following years and were upgraded to class C in the 2009 inventory.. All Figure 4.5: Walnut butt-log grade distribution in both 2006 and 2009 according to Canesin’s classification: thinned plots (4), not-thinned plots (10), and all plots (14). Plot 51.2 was not considered due to the lack of 2006 data.. Against any expectation, the amount of A class butt-logs drastically decreased also in the thinned plantations (from 18 down to 3 %). On the other hand B class trunks increased slightly (from 3 to 5 %), while D class dropped substantially from 65 to 47 % (Figure 4.5). This unexpected result could be due to the following reasons: • The thinned plots (four) are located near the river banks on soils with low fertility and high percentage of pebbles; therefore, the trees are not facing the best growing conditions and the tree architecture has not developed naturally into the hypotetical valuable shape;. 34.

(35) • They belong to the same owner and a similar management was applied: in particular, pruning operation have not been carried out in the proper way and at the right time; the small branches that were not recorded in 2006 grew bigger and took over the treshold of 3 cm considered in Canesin’s classification; some of them were then pruned too late. Consequently, several knots and branches with diameter larger than 3 cm were recorded in 2009 and most of the previous A class butt-logs fell in lower quality classes; • The subgective interpretation of both deviation and sinuosity could have caused several butt-logs to fall directly in C class. In fact, this is one of the limits of Canesin’s classification. Nosenzo’s classification solved partially this issue assigning different classes for the stem straightness and defining properly how to estimate the stem deviation from the straightness, the inclination and sinuosity.. 4.2. Wild cherry (Prunus avium L.). The stand parameters of all plots having wild cherry as principal tree species are reported in Table 4.5. The highest values of mean DBH (14 cm), mean tree height (10.1 m), mean DBH annual increment (1.55 cm/y) and mean height annual increment (1.21 m/y) in 2009 were recorded in the 9 years old plots. The lowest values were found in the oldest plots instead. In the winter 2007/2008, thinning operations were carried out in plots 2.1, 2.2, 16.1, and 16.2 and some cherry trees were felled especially in the first two plantations; mostly damaged and weaker trees were felled. As a result, some trees different from those sampled in 2006 had to be measured to reach the fixed number of 30 trees per each principal species in each plot and this could have affected slightly the 2009 stand parameters (because of the less contribution of smaller diameters). Therefore, the comparison with the 2006 data should be carefully interpreted in thinned plots. The DBH current annual increment (CAI) is generally in line with or even higher than the mean annual increment, especially in the youngest plantations where it reaches almost 1.5 cm/y. This suggests that the relatively young cherry trees have grown significantly during the last three years. Only in one case (plot 2.2) it was obtained an unexpected negative value for the DBH CAI. This could be due to the influence of the thinning operation when several cherry trees affected by diseases (especially rot) were. 35.

(36) removed in favour of the near ash trees. Mostly small diameters cherries remained and the mean DBH has decreased in 2009. The same explanation could be given also to justify the negative value of the height CAI in plot 2.1 and 2.2. Table 4.5: Mean DBH, mean butt-log height (stem H), and mean tree height (H) for all wild cherry trees both in 2006 and 2009 divided in age classes. The last four columns report the mean DBH and height annual increment (Im) and the mean DBH and height current annual increment (CAI, period 2006-2009). 2006 DBH H (cm) (m) 8.83 10.5 7.23 8.5 9.5 8.03. DBH (cm) 12.7 8.1 10.4. 2009 Stem H (m) 3.26 3.00 3.13. H (m) 8.63 7.12 7.83. dbh (cm/y) 0.91 0.58 0.74. h (m/y) 0.62 0.51 0.56. 8.8. 6.75. 10.7. 2.61. 8.58. 0.89. 0.71. 0.66. 0.61. 30 30 30 90. 5.1 9.1 9.1 7.8. 6.25 7.12 6.92 6.8. 6.0 11.0 10.0 9.03. 2.85 2.61 2.92 2.79. 6.22 7.56 7.50 7.15. 0.50 1.00 0.91 0.82. 0.52 0.69 0.68 0.65. 0.30 0.64 0.31 0.42. 0.00 0.15 0.19 0.13. 10 10 10. 30 30 60. 9.5 8.3 8.9. 7.13 5.85 6.49. 12.1 13.8 12.9. 3.95 2.65 3.30. 8.46 7.26 7.85. 1.21 1.38 1.29. 0.85 0.73 0.78. 0.86 1.83 1.34. 0.44 0.47 0.45. 35 37A 41.1 41.2 51.1 51.2 53 54A Mean. 9 9 9 9 9 9 9 9 9. 30 30 30 30 30 30 30 30 240. 10.7 10.2 9.4 9.0 9.2 11.5 6.5 9.5. 7.82 8.23 6.42 6.03 7.50 7.20 6.43 7.09. 15.9 13.8 14.8 15.0 13.0 13.8 16.2 9.4 14.0. 4.46 3.09 2.30 2.12 2.88 2.88 2.59 2.64 2.87. 10.68 10.04 8.43 9.03 12.04 12.04 9.28 6.82 10.10. 1.77 1.53 1.64 1.67 1.45 1.53 1.80 1.05 1.55. 1.19 1.12 0.94 1.00 1.34 1.34 1.03 0.76 1.21. 1.72 1.20 1.79 2.00 1.26 1.58 0.98 1.49. 0.95 0.61 0.67 1.00 1.52 0.69 0.13 1.00. 61 77A.1 77A.2 Mean. 8 8 8 8. 30 30 30 90. 6.7 7.3 4.5 6.2. 5.50 6.77 4.91 5.73. 12.8 11.9 7.1 10.6. 2.95 2.86 2.75 2.85. 7.25 10.21 6.69 8.60. 1.60 1.48 0.89 1.32. 0.91 1.28 0.84 1.08. 2.03 1.51 0.88 1.48. 0.58 1.15 0.59 0.96. Plot n° 2.1 2.2 Mean. Age (y) 14 14 14. 12A. 12. 30. 15A 16.1 16.2 Mean. 11 11 11 11. 22 30 Mean. N° 30 30 60. Im. CAI dbh h (cm/y) (m/y) -0.07 0.75 -0.04 -0.12 0.75. Cherry trees manifest a clear reduction of mean DBH annual increment growing older. This phenomenon is shown by the straight curve in Figure 4.6 where the mean DBH annual increment is plotted against the age. The graph suggests that competition starts when the trees are 10 years old and since then it increases causing growth increment reduction. Most of the points to the right represent the thinned plantations; therefore, it is likely that those points will shift slightly upwards during the next years as more light has become available to the remaining cherry trees.. 36.

(37) Figure 4.6: Mean DBH annual increment of cherry trees plotted against age.. The amount of stem defects recorded for cherry trees was less than that for walnut trees and the sample size is also higher (570 trees); as a result the average number of defects per each cherry butt-log (3 defects/trunk) is less than that found for walnut (4 defects/trunk). The two stem defects groups dealing with the presence of knots and branches and unsuitable stem shape are the most relevant (54 and 38 % of total defects respectively) (Figure 4.7).. Figure 4.7: Stem defects found on all wild cherry stems. See Table 4.1 for the list of stem defects included in each group.. Inside the defect group “knots and branches”, the most common are covered knots (37%), sum of branches and knots diameters more than 60 mm (28 %) and branches and knots with diameter larger than 3 cm (16 %).. 37.

(38) It should be noticed the higher percentage of the defect “sum of the diameters of branches and knots between 15 and 60 mm” (7 %) compared to the value found in walnut trees (2 %); this and the inferior stem shape defects are the main reasons that allow more cherry butt-logs to fall in “B class” according to Nosenzo’s classification or even in A class when adopting Canesin’s classification (Figure 4.8).. Figure 4.8: Stem defects due to the presence of knots and branches on all wild cherry butt-logs.. The detailed results about stem quality in each sample plot having wild cherry as principal species is reported in Table 4.6, while the total share of A, B, C and D classes for all walnut trees is reported in Figure 4.9. As in the walnut stem quality analysis, Canesin’s classification produces higher percentage of the higher quality classes (A and B) when compared to Nosenzo’s classification. The Fisher exact test was applied to investigate the statistical difference between the number of stems that fall in each quality class with Nosenzo’s classification and the number of stems in each quality class with Canesin’s classification. The test resulted to be very significant in 63 % of the cases (12 plots out of 19), and not significant in only 21 % of the cases (4 plots out of 19). Therefore, the two stem classifications can be considered significantly different when applied on cherry stems. The total amount of “D class” is similar and lower than that found in walnut trees, but the disparities between the two stem classifications adopted are relevant for the other quality classes. Very significant is the different percentage of butt-logs that fall in “A class”: 3% (Nosenzo’s) versus 35 % (Canesin’s).. 38.

(39) Table 4.6: Wild cherry butt-log assortments obtained with Nosenzo’s and Canesin’s classification in the 19 sample plots having wild cherry as principal species. The last two columns show the p value and the relative significance level obtained with the Fisher exact test (*=little significant; **=significant; ***=very significant, n.s.=not significant) Nosenzo's classification. Canesin's classification. Plot n°. Age (y). A. B. C. D. A. B. C. D. p (Fisher). Significance. 2.1. 14. 3%. 3%. 80%. 13%. 27%. 37%. 23%. 13%. 0.000. ***. 2.2. 14. 0%. 17%. 30%. 53%. 17%. 10%. 20%. 53%. 0.104. n.s.. 12A. 12. 0%. 10%. 53%. 37%. 13%. 3%. 47%. 37%. 0.206. n.s.. 15A. 12. 10%. 23%. 40%. 27%. 30%. 3%. 40%. 27%. 0.057. *. 16.1. 11. 10%. 17%. 47%. 27%. 33%. 27%. 13%. 27%. 0.018. **. 16.2. 11. 0%. 20%. 47%. 33%. 20%. 27%. 20%. 33%. 0.021. **. 22. 10. 3%. 17%. 73%. 7%. 63%. 0%. 30%. 7%. 0.000. ***. 30. 10. 0%. 33%. 40%. 27%. 53%. 0%. 20%. 27%. 0.000. ***. 35. 9. 0%. 7%. 83%. 10%. 27%. 40%. 27%. 7%. 0.000. ***. 37A. 9. 0%. 37%. 47%. 17%. 47%. 3%. 33%. 17%. 0.000. ***. 41.1. 9. 0%. 7%. 30%. 63%. 13%. 3%. 20%. 63%. 0.165. n.s.. 41.2. 9. 0%. 3%. 20%. 77%. 3%. 10%. 10%. 77%. 0.404. n.s.. 51.1. 9. 0%. 20%. 70%. 10%. 47%. 3%. 40%. 10%. 0.000. ***. 51.2. 9. 17%. 33%. 40%. 10%. 53%. 3%. 33%. 10%. 0.002. ***. 53. 9. 0%. 10%. 50%. 40%. 43%. 3%. 13%. 40%. 0.000. ***. 54A. 9. 10%. 30%. 30%. 30%. 40%. 0%. 30%. 30%. 0.001. ***. 61. 8. 0%. 13%. 70%. 17%. 20%. 47%. 17%. 17%. 0.000. ***. 77A.1. 8. 0%. 57%. 33%. 10%. 60%. 7%. 23%. 10%. 0.000. ***. 77A.2. 8. 3%. 47%. 33%. 17%. 53%. 0%. 33%. 13%. 0.000. ***. Figure 4.9: Wild cherry butt-log grade distribution according to Nosenzo’s classification (left) and Canesin’s classification (right).. However, Nosenzo’s classification demonstrated to be more precise and objective; therefore it will be used to assess the 2009 stand quality; looking at the results, A and B classes together are more than 20 % of the total in 11 plots out of 19: the maximum amount of high quality timber is found in plot 77A.1 (57 %, but only B class) followed 39.

(40) by plot 77A.2 and 51.2 (50 %, of which 17 % of only A class). If this is a better result compared to walnut trunks, the economic benefits will be quite low and questionable in 40 % of the plantations having cherry as principal species. The Fisher exact test applied to compare the 2006 butt-log grade distribution with the corresponding 2009 quality classes obtained with Canesin’s classification resulted very significant in 61 % of the cases (11 plots out of 18) and not significant in 28 % of the cases (5 plots out of 18) (Table 4.7). Plot 51.2 was excluded from this analysis because of the lack of 2006 data. Also the differences between the total number of stems in each quality class found in 2006 and 2009 resulted to be very significant (X2= 69.1; p value = 0.000); only “A class” remained the same (34 %), while the other three classes changed significantly: “B class” decreased slightly, while “C class” has triplicated (from 8 to 26%) and “D class” decreased significantly (from 41 to 28 %) (Figure 4.10). Table 4.7: Wild cherry butt-log assortments obtained in 2006 and 2009 with Canesin’s classification in the 19 sample plots having wild cherry as principal species. The last two columns show the p value and the equivalent significance level obtained with the Fisher exact test (*=little significant; **=significant; ***=very significant, n.s.=not significant) 2006 (Canesin’s class.). 2009 (Canesin's class.) A. B. C. D. p (Fisher). Significance. Plot n°. Age (y). A. 2.1. 14. 73%. 7%. 13%. 7%. 27%. 37%. 23%. 13%. 0.002. ***. 2.2. 14. 33%. 10%. 7%. 50%. 17%. 10%. 20%. 53%. 0.303. n.s.. 12A. 12. 30%. 23%. 10%. 37%. 13%. 3%. 47%. 37%. 0.003. ***. 15A. 12. 80%. 7%. 3%. 10%. 30%. 3%. 40%. 27%. 0.000. ***. 16.1. 11. 27%. 23%. 10%. 40%. 33%. 27%. 13%. 27%. 0.778. n.s.. 16.2. 11. 40%. 10%. 0%. 50%. 20%. 27%. 20%. 33%. 0.009. ***. 22. 10. 50%. 27%. 10%. 13%. 63%. 0%. 30%. 7%. 0.003. ***. 30. 10. 43%. 23%. 7%. 27%. 53%. 0%. 20%. 27%. 0.022. **. 35. 9. 17%. 37%. 23%. 23%. 27%. 40%. 27%. 7%. 0.322. n.s.. 37A. 9. 20%. 33%. 17%. 30%. 47%. 3%. 33%. 17%. 0.003. ***. 41.1. 9. 10%. 10%. 3%. 77%. 13%. 3%. 20%. 63%. 0.182. n.s.. 41.2. 9. 10%. 7%. 3%. 80%. 3%. 10%. 10%. 77%. 0.605. n.s.. 51.1. 9. 30%. 17%. 0%. 53%. 47%. 3%. 40%. 10%. 0.000. ***. 51.2. 9. -. -. -. 53%. 3%. 33%. 10%. -. -. 53. 9. 13%. 40%. 7%. 40%. 43%. 3%. 13%. 40%. 0.001. ***. 54A. 9. 27%. 13%. 0%. 60%. 40%. 0%. 30%. 30%. 0.000. ***. 61. 8. 20%. 7%. 10%. 63%. 20%. 47%. 17%. 17%. 0.000. ***. 77A.1. 8. 23%. 20%. 10%. 47%. 60%. 7%. 23%. 10%. 0.001. ***. 77A.2. 8. 57%. 7%. 7%. 30%. 53%. 0%. 33%. 13%. 0.021. **. 40. B. C. -. D.

(41) The Chi-square test became very significant also when comparing the 2006 and the 2009 values of the thinned plantations (X2 = 18.1; p value = 0.000). In this case, all classes change significantly: A and D classes decreases, while B and C classes increases. Probably, most of the damaged trees that were considered D class in 2006 were removed by the thinning operation and this could explain the reduction of buttlogs in the lowest class. Nevertheless, the still high percentage of D class (32 %) is due to the high amount of very weak and dominated trees (graded as D) recorded in 2009. The strong reduction of butt-logs in the A class (from 43 to 24 %) is mainly due to the inappropriate and late pruning operations (all thinned plantations belong to the same owner and saw a similar management); branches less than 3 cm in 2006 grew larger and most of them were pruned in recent years leaving fresh knots with diameters larger than 3 cm. Therefore, the presence of only one knot > 3 cm has downgraded the whole buttlog in B class or even in C class when both branches and knots > 3 cm are present.. All Figure 4.10: Wild cherry butt-log grade distribution in both 2006 and 2009 according to Canesin’s classification: thinned plots (4), not-thinned plots (14), and all plots (18). Plot 51.2 was not considered due to the lack of 2006 raw data.. An additional sign of improper management is the presence of several epicormic shoots that were produced after the thinning operation when more light reached the ground. To avoid stem downgrading, epicormic shoots should have been pruned immediately after their appearance but this has not happened. However, their presence on the butt-logs affected mostly Nosenzo’s classification; in fact, even 6 or 7 epicormic shoots with an. 41.